A Study of Automatic Defects Extraction of X-ray Weld Image Based on Computed Radiography System

According to the characteristic of X-ray images of weld, defect extraction techniques were studied. Firstly, on the basis of wavelet analysis, a new wavelet adaptive threshold de-noising method based on genetic algorithm optimization is proposed. Secondly, an algorithm of multi-scale morphological to local contrast enhancement is designed. Finally, background is simulated, and the defect regions were extracted using algorithm of digital subtraction. We can accomplish defect extraction by image segmentation. The experimental results indicate that the methods can achieve automatic extraction of defect region, which lays a good foundation for flaw feature parameter extraction and choice.